Box Suite Recommendation

Stuart Rogers

DOI: https://doi.org/10.48550/arXiv.2004.11533

2021-01-03

Abstract:An algorithm for recommending a suite of boxes for shipping a retailer's online customer orders is presented.

Optimization and Control

What problem does this paper attempt to address?

BOXREC: Recommending a Box of Preferred Outfits in Online Shopping

Debopriyo Banerjee,Krothapalli Sreenivasa Rao,Shamik Sural,Niloy Ganguly

DOI: https://doi.org/10.1145/3408890

2024-02-26

Abstract:Over the past few years, automation of outfit composition has gained much attention from the research community. Most of the existing outfit recommendation systems focus on pairwise item compatibility prediction (using visual and text features) to score an outfit combination having several items, followed by recommendation of top-n outfits or a capsule wardrobe having a collection of outfits based on user's fashion taste. However, none of these consider user's preference of price-range for individual clothing types or an overall shopping budget for a set of items. In this paper, we propose a box recommendation framework - BOXREC - which at first, collects user preferences across different item types (namely, top-wear, bottom-wear and foot-wear) including price-range of each type and a maximum shopping budget for a particular shopping session. It then generates a set of preferred outfits by retrieving all types of preferred items from the database (according to user specified preferences including price-ranges), creates all possible combinations of three preferred items (belonging to distinct item types) and verifies each combination using an outfit scoring framework - BOXREC-OSF. Finally, it provides a box full of fashion items, such that different combinations of the items maximize the number of outfits suitable for an occasion while satisfying maximum shopping budget. Empirical results show superior performance of BOXREC-OSF over the baseline methods.

Information Retrieval
A Note on Online Strip Packing

Deshi Ye,Xin Han,Guochuan Zhang

DOI: https://doi.org/10.1007/s10878-007-9125-x

2008-01-01

Journal of Combinatorial Optimization

Abstract:In online strip packing we are asked to pack a list of rectangles one by one into a vertical strip of unit width, without any information about future rectangles. The goal is to minimize the total height of strip used. The best known algorithm is First Fit Shelf algorithm (Baker and Schwarz in SIAM J. Comput. 12(3):508–525, 1983), which has an absolute competitive ratio of 6.99 under the assumption that the height of each rectangle is bounded from above by one. We improve the shelf algorithm and show an absolute competitive ratio of $7/2+\sqrt{10}\approx 6.6623$ without the restriction on rectangle heights. Our algorithm also beats the best known online algorithm for parallel job scheduling.
Online Algorithms for 1-Space Bounded 2-Dimensional Bin Packing and Square Packing

Yong Zhang,Francis Y. L. Chin,Hing-Fung Ting,Xin Han,Chung Keung Poon,Yung H. Tsin,Deshi Ye

DOI: https://doi.org/10.1016/j.tcs.2014.02.045

IF: 1.002

2014-01-01

Theoretical Computer Science

Abstract:In this paper, we study 1-space bounded 2-dimensional bin packing and square packing. A sequence of rectangular items (square items) arrive one by one, each item must be packed into a square bin of unit size on its arrival without any information about future items. When packing items, 90°-rotation is allowed. 1-space bounded means there is only one “active” bin. If the “active” bin cannot accommodate the coming item, it will be closed and a new bin will be opened. The objective is to minimize the total number of bins used for packing all items in the sequence.
An Asymptotic Competitive Scheme for Online Bin Packing.

Lin Chen,Deshi Ye,Guochuan Zhang

DOI: https://doi.org/10.1016/j.tcs.2015.04.038

IF: 1.002

2015-01-01

Theoretical Computer Science

Abstract:In the online bin packing problem, a list of items with integral sizes between 1 to B arrive one by one. Each item must be irrevocably assigned into a bin of capacity B upon its arrival without any information on the subsequent items, and the goal is to minimize the number of used bins. In this paper, we deal with online bin packing from a new sight. We present an asymptotic competitive scheme, i.e., for any ϵ>0, the asymptotic competitive ratio is at most ρ⁎+ϵ, where ρ⁎ is the smallest possible asymptotic competitive ratio among all online algorithms. Apart from the technical results, the analysis to bridge the online and the off-line approaches might be of particular interests.
Online Removable Square Packing

Xin Han,Kazuo Iwama,Guochuan Zhang

DOI: https://doi.org/10.1007/s00224-007-9039-0

IF: 0.616

2008-01-01

Theory of Computing Systems

Abstract:The online removable square packing problem is a two-dimen-sional version of the online removable Knapsack problem. For a sequence of squares with side length at most 1, we are requested to pack a subset of them into a unit square bin in an online fashion where the online player can decide whether to take the current square or not and which squares currently in the unit square to remove. The goal is to maximize the total packed area. Our results include: (i) No online algorithm can achieve a better competitive ratio than $(\sqrt{5}+3)/2\approx 2.618$. (ii) The matching upper bound is achieved by a relatively simple online algorithm if repacking is allowed. (iii) Without repacking, we can achieve an upper bound of 3 by using the concept of bricks of Januszewski and Lassak. (iv) The offline version of the problem admits a PTAS.
A Machine Learning Approach to Shipping Box Design

Guang Yang,Cun Mu

DOI: https://doi.org/10.1007/978-3-030-29516-5_32

2019-08-24

Abstract:Having the right assortment of shipping boxes in the fulfillment warehouse to pack and ship customer’s online orders is an indispensable and integral part of nowadays eCommerce business, as it will not only help maintain a profitable business but also create great experiences for customers. However, it is an extremely challenging operations task to strategically select the best combination of tens of box sizes from thousands of feasible ones to be responsible for hundreds of thousands of orders daily placed on millions of inventory products. In this paper, we present a machine learning approach to tackle the task by formulating the box design problem prescriptively as a generalized version of weighted k-medoids clustering problem, where the parameters are estimated through a variety of descriptive analytics. We test this machine learning approach on fulfillment data collected from Walmart U.S. eCommerce, and our approach is shown to be capable of improving the box utilization rate by more than $$10\%$$.
A decision-tree framework to select optimal box-sizes for product shipments

Karthik S. Gurumoorthy,Abhiraj Hinge

DOI: https://doi.org/10.48550/arXiv.2202.04277

IF: 5.414

2022-02-09

Machine Learning

Abstract:In package-handling facilities, boxes of varying sizes are used to ship products. Improperly sized boxes with box dimensions much larger than the product dimensions create wastage and unduly increase the shipping costs. Since it is infeasible to make unique, tailor-made boxes for each of the $N$ products, the fundamental question that confronts e-commerce companies is: How many $K << N$ cuboidal boxes need to manufactured and what should be their dimensions? In this paper, we propose a solution for the single-count shipment containing one product per box in two steps: (i) reduce it to a clustering problem in the $3$ dimensional space of length, width and height where each cluster corresponds to the group of products that will be shipped in a particular size variant, and (ii) present an efficient forward-backward decision tree based clustering method with low computational complexity on $N$ and $K$ to obtain these $K$ clusters and corresponding box dimensions. Our algorithm has multiple constituent parts, each specifically designed to achieve a high-quality clustering solution. As our method generates clusters in an incremental fashion without discarding the present solution, adding or deleting a size variant is as simple as stopping the backward pass early or executing it for one more iteration. We tested the efficacy of our approach by simulating actual single-count shipments that were transported during a month by Amazon using the proposed box dimensions. Even by just modifying the existing box dimensions and not adding a new size variant, we achieved a reduction of $4.4\%$ in the shipment volume, contributing to the decrease in non-utilized, air volume space by $2.2\%$. The reduction in shipment volume and air volume improved significantly to $10.3\%$ and $6.1\%$ when we introduced $4$ additional boxes.
Inventory Based Recommendation Algorithms

Du Chen,Yuming Deng,Guangrui Ma,Hao Ge,Yunwei Qi,Ying Rong,Xun Zhang,Huan Zheng

DOI: https://doi.org/10.1109/bigdata50022.2020.9378261

2020-01-01

Abstract:We propose two recommendation algorithms for e-commerce with supply limits, a scenario that has not been intensively studied in the literature. One algorithm is a linear programming-based algorithm that uses historical data to approximate customer arrival patterns and generate shadow prices for inventories. The price of inventory can be introduced to existing recommendation algorithms to obtain adjusted rankings for recommendation. The other algorithm balances expected revenue and inventory consumption, and it uses a simple penalty function to reduce the chance of recommending low-inventory-level products. Both algorithms are suitable for online recommendation systems for grocery stores with both online and offline channels, and can incorporate the features of perishable products, which need to be sold within limited time. Both algorithms are tested in a simulation using estimated parameters from Freshippo, a supermarket owned by the Alibaba Group. The numerical results show that both algorithms can generate higher sales volume and higher revenue.
EXPRESS: A Machine Learning Approach to Solve the E-commerce Box-Sizing Problem

Shanthan Kandula,Debjit Roy,Kerem Akartunalı

DOI: https://doi.org/10.1177/10591478241282249

IF: 4.638

2024-09-01

Production and Operations Management

Abstract:E-commerce packages are notorious for their inefficient usage of space. More than one-quarter volume of a typical e-commerce package comprises air and filler material. The inefficient usage of space significantly reduces the transportation and distribution capacity increasing the operational costs. Therefore, designing an optimal set of packaging box sizes is crucial for improving efficiency. We present the first learning-based framework to determine the optimal packaging box sizes. In particular, we propose a three-stage optimization framework that combines unsupervised learning, reinforcement learning, and tree search to design box sizes. The package optimization problem is formulated into a sequential decision-making task called the box-sizing game. A neural network agent is then designed to play the game and learn heuristic rules to solve the problem. In addition, a tree-search operator is developed to improve the performance of the learned networks. When benchmarked with company-based optimization formulation and two alternate optimization models, we find that our ML-based approach can effectively solve large-scale problems within a stipulated time. We evaluated our model on real-world datasets supplied by a large e-commerce platform. The framework is currently adopted by a large e-commerce company across its 28 fulfillment centers, which is estimated to save the company about 7.1 million USD annually. In addition, it is estimated that paper consumption will be reduced by 2080 metric tons and greenhouse gas emissions by 1960 metric tons annually. The presented optimization framework serves as a decision support tool for designing packaging boxes at large e-commerce warehouses.

operations research & management science,engineering, manufacturing
Build Your Own Bundle - A Neural Combinatorial Optimization Method

Qilin Deng,Kai Wang,Minghao Zhao,Runze Wu,Yu Ding,Zhene Zou,Yue Shang,Jianrong Tao,Changjie Fan

DOI: https://doi.org/10.1145/3474085.3475440

2021-01-01

Abstract:In the business domain, bundling is one of the most important marketing strategies to conduct product promotions, which is commonly used in online e-commerce and offline retailers. Existing recommender systems mostly focus on recommending individual items that users may be interested in, such as the considerable research work on collaborative filtering that directly models the interaction between users and items. In this paper, we target at a practical but less explored recommendation problem named personalized bundle composition, which aims to offer an optimal bundle (i.e., a combination of items) to the target user. To tackle this specific recommendation problem, we formalize it as a combinatorial optimization problem on a set of candidate items and solve it within a neural combinatorial optimization framework. Extensive experiments on public datasets are conducted to demonstrate the superiority of the proposed method.
Multiple bin-size bin packing problem considering incompatible product categories

Yu-Chung Tsao,Jo-Ying Tai,Thuy-Linh Vu,Tsung-Hui Chen

DOI: https://doi.org/10.1016/j.eswa.2024.123340

IF: 8.5

2024-01-31

Expert Systems with Applications

Abstract:Optimizing the order-fulfillment process is crucial to the logistics industry. Replacing experience-based judgment with intelligent methods in the packing process not only improves work efficiency but also eliminates resource wastage. This study focuses on the three-dimensional multiple bin-size bin packing problem with compatible categories (3D-MBSBPPCC) and different size of the boxes. An algorithm is developed to achieve the best packing pattern while minimizing the number of boxes required and the total unused space in the boxes. In the algorithm, we first use the first-fit decreasing algorithm to find an initial solution. Next, the carton configuration algorithm is applied to perform packing and evaluate the solution. Subsequently, a simulated annealing (SA)-based algorithm is employed to improve the solution by moving the products among the subsets of the solution. Finally, hybrid metaheuristics based on both SA and the genetic algorithm , are proposed to improve the solution quality. We tested the proposed packing algorithms on real-world scenarios. The experimental results demonstrate that the proposed algorithms are effective in solving 3D-MBSBPPCC.

computer science, artificial intelligence,engineering, electrical & electronic,operations research & management science
Order Effects in Online Product Recommendation: A Scenario-based Analysis.

Xunhua Guo,Mingyue Zhang,Chenyue Yang,Guoqing Chen

2016-01-01

Abstract:Existing research on recommender systems has mostly focused on developing algorithms to predict the degrees to which a consumer may like a product. Recommended products are usually presented based on their predicted ratings in descending order. Considering the cognitive process of online consumers, when users evaluate the recommended products in sequence, the descending order may not necessarily be a good solution to increase consumers' willingness to purchase. Drawing upon the evaluability perspective and the order effects theory, we formulate a scenario where each product has two attributes, one of which can be evaluated independently while the other is difficult to evaluate without comparison. Analysis shows that in two out of the three cases of the scenario, presenting the most recommended product in the second place will result in higher consumer willingness to pay. The findings provide a new angle for understanding the behavioral implications of using recommender systems in ecommerce.
For the Shipbuilding Enterprise Model of Intelligent Recommendation Services

Zhang Wei,Pan Xiao Hong,Fang Xin Kang

DOI: https://doi.org/10.4028/www.scientific.net/amm.34-35.222

2010-01-01

Applied Mechanics and Materials

Abstract:Based on ships transaction in useful information personalization gain question, proposed that one kind the intelligent recommendation service model which unifies the ASP (Application Service Providers) pattern and the recommendation technology. In this model, the service uses five structures, the intelligence management transfer application service level and the resources component level, cooperates to realize the intelligence mutually. In view of the ships profession electronic commerce pattern's characteristic, the depth limited auto-adapted k close neighbor searching algorithm which uses transforms the user grading non-isotropic space as the isotropic space, obtains the isotropic grading matrix, thus searches the current user k recent neighbor, has current user's forecast grading, and has the recommendation. Has provided the simple direct-viewing practical commercial service for the ships manufacturing firm.
Heuristic algorithms based on column generation for an online product shipping problem

Wei Wu,Mayu Ito,Yannan Hu,Hiromichi Goko,Mihiro Sasaki,Mutsunori Yagiura

DOI: https://doi.org/10.1016/j.cor.2023.106403

2023-09-02

Abstract:In this paper, we consider a product shipping problem (PSP). Given a time horizon and a set of products, each with a weight, a release date, and a due date, the PSP calls for product shipping to be planned between a fixed origin–destination pair so as to minimize the total shipping cost. Given several types of boxes with different sizes and shipping costs, packing several products into one larger box may lead to cheaper shipping costs than sending them separately in smaller boxes. The online PSP is a problem with an online constraint that requires a decision to be made for each day without knowing about future demands or being able to change boxes that have already been packed and shipped. For the online PSP, we propose column-generation-based algorithms with six different criteria, incorporating ideas such as iterative surplus reduction, and heuristic column generation. In the computational experiments, we tested 15 types of online algorithms under each criterion. Computational results on three different sets of instances show that one type of the proposed algorithms with one of the six criteria called "regret with cost of remaining capacity" obtains high-quality solutions with average gaps of 3.4% over all the tested instances.

computer science, interdisciplinary applications,engineering, industrial,operations research & management science
Think out of the package: Recommending package types for e-commerce shipments

Karthik S. Gurumoorthy,Subhajit Sanyal,Vineet Chaoji

DOI: https://doi.org/10.48550/arXiv.2006.03239

2020-06-05

Abstract:Multiple product attributes like dimensions, weight, fragility, liquid content etc. determine the package type used by e-commerce companies to ship products. Sub-optimal package types lead to damaged shipments, incurring huge damage related costs and adversely impacting the company's reputation for safe delivery. Items can be shipped in more protective packages to reduce damage costs, however this increases the shipment costs due to expensive packaging and higher transportation costs. In this work, we propose a multi-stage approach that trades-off between shipment and damage costs for each product, and accurately assigns the optimal package type using a scalable, computationally efficient linear time algorithm. A simple binary search algorithm is presented to find the hyper-parameter that balances between the shipment and damage costs. Our approach when applied to choosing package type for Amazon shipments, leads to significant cost savings of tens of millions of dollars in emerging marketplaces, by decreasing both the overall shipment cost and the number of in-transit damages. Our algorithm is live and deployed in the production system where, package types for more than 130,000 products have been modified based on the model's recommendation, realizing a reduction in damage rate of 24%.

Machine Learning,Optimization and Control
Recommending Short-lived Dynamic Packages for Golf Booking Services

Robin Swezey,Young-joo Chung

DOI: https://doi.org/10.48550/arXiv.2103.07779

2021-03-13

Information Retrieval

Abstract:We introduce an approach to recommending short-lived dynamic packages for golf booking services. Two challenges are addressed in this work. The first is the short life of the items, which puts the system in a state of a permanent cold start. The second is the uninformative nature of the package attributes, which makes clustering or figuring latent packages challenging. Although such settings are fairly pervasive, they have not been studied in traditional recommendation research, and there is thus a call for original approaches for recommender systems. In this paper, we introduce a hybrid method that leverages user analysis and its relation to the packages, as well as package pricing and environmental analysis, and traditional collaborative filtering. The proposed approach achieved appreciable improvement in precision compared with baselines.
Optimal Recommendation in the Presense of Comparison Shopping

Hai Jiang

2012-01-01

Abstract:Recommender systems have been widely used by online stores to suggest items of interest to users. These systems often identify a subset of items from a much larger set that best matches the user’s interest. A key concern with existing approaches is over-specialization, which results in returning items that are too similar to each other. Unlike existing solutions that rely on diversity metrics to reduce similarity among recommended items, we propose to use choice probability to measure the overall quality of recommendation lists, which unifies the desire to achieve both relevancy and diversity in generating recommendations. We first define the recommendation problem in the discrete choice setting. We then develop an efficient greedy algorithm that solves the problem to optimum in O(KlogK) time, where K stands for the number of wanted items. Using data from a major travel website, we show that recommendation lists generated by our algorithm outperform existing strong baselines. Sensitivity analysis is also conducted to evaluate the robustness of our solution.
Supplier Recommendation in Online Procurement

Victor Coscrato,Derek Bridge

2024-03-03

Abstract:Supply chain optimization is key to a healthy and profitable business. Many companies use online procurement systems to agree contracts with suppliers. It is vital that the most competitive suppliers are invited to bid for such contracts. In this work, we propose a recommender system to assist with supplier discovery in road freight online procurement. Our system is able to provide personalized supplier recommendations, taking into account customer needs and preferences. This is a novel application of recommender systems, calling for design choices that fit the unique requirements of online procurement. Our preliminary results, using real-world data, are promising.

Information Retrieval,Machine Learning
Integrated Bundling and Pricing of Unique Items

Maxime Bouscary,Mazen Danaf,Saurabh Amin

2024-09-05

Abstract:Retailers have significant potential to improve recommendations through strategic bundling and pricing. By taking into account different types of customers and their purchasing decisions, retailers can better accommodate customer preferences and increase revenues while reducing unsold items. We consider a retailer seeking to maximize its expected revenue by selling unique and non-replenishable items over a finite horizon. The retailer may offer each item individually or as part of a bundle. Our approach provides tractable bounds on expected revenue that are tailored to unique items and suitable for a rich class of choice models. We leverage these bounds to propose a bundling algorithm that efficiently selects bundles in a column-generation fashion. Under the multinomial logit model, our bounds are asymptotically optimal as the expected number of arrivals $\lambda$ grows, yielding a performance bound in $O(1/\lambda)$. In contrast, we show that both the static and fluid approximations are not asymptotically optimal. Moreover, we propose a greedy algorithm that allows for tractable dynamic bundling. We show through numerical experiments that third-party logistics providers (3PL) in particular can benefit from improved load recommendations and pricing decisions. By not only minimizing shipping costs, but also reducing empty miles and suggesting options that better match carrier preferences, our methodology benefits digital brokerage platforms while contributing to the transportation industry's sustainability efforts. In a Texas Triangle simulation powered by Uber Freight data, dynamic bundling and pricing reduces costs by 6%, empty miles by 25%, and more than halves the number of unmatched loads over dynamic pricing alone.

Optimization and Control
A comprehensive solution strategy for right-sized packaging

John C. Peck,Edward W. Page,James A. Bate

DOI: https://doi.org/10.69554/czup3564

2023-03-01

Abstract:Small parcel carriers are now basing shipping charges primarily on the dimensions of shipping cartons rather than the weight of packed cartons. Consequently, reducing wasted space in cartons can result in significant savings in shipping costs as well as corrugated material and dunnage costs. Most companies maintain a relatively small number of carton sizes, so the opportunity to select a carton to contain order items with a small amount of empty space is limited. Furthermore, choosing the best carton to pack items, either before or after they have been picked, is difficult for packers using the ‘eyeball’ method of selection and often requires repacking, which further adds to costs. Maintaining a large number of different carton sizes (eg 40+) increases the likelihood that a carton selected for packing will have minimal wasted space. This paper describes the challenges and artificial intelligence (AI)-based solutions developed to enable the determination of which carton sizes to maintain in a large set of different carton sizes, automatically calculate the best carton(s) for packing an order from the large set, quickly identify the best carton for a packer, efficiently replenish consumed cartons from a carton supply area, and simplify vendor managed inventory (VMI) by a carton supplier to enable frequent delivery of replenishment cartons.

Box Suite Recommendation

BOXREC: Recommending a Box of Preferred Outfits in Online Shopping

A Note on Online Strip Packing

Online Algorithms for 1-Space Bounded 2-Dimensional Bin Packing and Square Packing

An Asymptotic Competitive Scheme for Online Bin Packing.

Online Removable Square Packing

A Machine Learning Approach to Shipping Box Design

A decision-tree framework to select optimal box-sizes for product shipments

Inventory Based Recommendation Algorithms

EXPRESS: A Machine Learning Approach to Solve the E-commerce Box-Sizing Problem

Build Your Own Bundle - A Neural Combinatorial Optimization Method

Multiple bin-size bin packing problem considering incompatible product categories

Order Effects in Online Product Recommendation: A Scenario-based Analysis.

For the Shipbuilding Enterprise Model of Intelligent Recommendation Services

Heuristic algorithms based on column generation for an online product shipping problem

Think out of the package: Recommending package types for e-commerce shipments

Recommending Short-lived Dynamic Packages for Golf Booking Services

Optimal Recommendation in the Presense of Comparison Shopping

Supplier Recommendation in Online Procurement

Integrated Bundling and Pricing of Unique Items

A comprehensive solution strategy for right-sized packaging